[Pipet Devel] ideas for a smaller piper core

J.W. Bizzaro jeff at bioinformatics.org
Mon May 28 10:45:20 EDT 2001


I've been thinking a bit about the direction that I'd like to see Piper head,
particularly in the short-term where our problem is making a usable system
right away (this will help attract users and more developers).

I think to solve the usability problem, we should look at ways to get Piper
compiled and running with less requirements.  And the requirements to consider
are the ones that are not for the core of Piper (the DL and the BL).  These
are the more elaborate UI's and PL's, such as Pied and Overflow.  NOW, I'M NOT
SAYING THAT WE SHOULD NOT SUPPORT THESE!  I'm just saying that they should not
be required to compile and run Piper.

Regarding Pied, I would like to distribute it as a separate package.  I've
mentioned this before.  I think that Peep, our console UI, should be the only
one distributed with the core of Piper.  This would remove the requirement of
the GTK/GNOME and PyGTK libraries (although Jarl still uses GTK in the BL).

I personally would like to see Piper used as a "better UNIX GUI", allowing CLI
progs to be connected via GUI.  It should be no surprise, since most of my
announcements about Piper have mentioned this.  Piper is, of course, a
dataflow system, but there aren't as many systems that do what I'd like Piper
to do with CLI progs, and I think that we'd get more users this way.  In light
of this, I'd like to rename Pied as "Piper Desktop".  (There are some other
Pied (Desktop) issues that I will address later.)

Regarding Overflow and the "better UNIX GUI" idea, I'd like to have a very
small PL distributed with the core of Piper that simply wraps CLI prorgams,
per my discourse:


YES, I know that Overflow does this and has a better way of doing it, and I'd
like to do it Overflow's way.  I simply want this feature by default in Piper
without requiring the use of a sophisticated data flow system.  As for myself,
I WOULD WANT to use Overflow, but perhaps many Piper users will not find the
need for it.

Also, I would like to see systems like SRS


and Narval plug directly into the BL as PL's.  In other words, we could have
any number of PL's.  We should produce some API docs for that.  As Jarl
mentioned to me, each PL, if not custom made for Piper, would require a small
wrapper.  (So, the code we now have in Piper for wrapping Overflow can also be

So, what are your thoughts on this?  Yes, no?

J.W. Bizzaro                                jeff at bioinformatics.org
Director, Bioinformatics.org        http://bioinformatics.org/~jeff
"As we enjoy great advantages from the inventions of others, we
should be glad of an opportunity to serve others by any invention
of ours; and this we should do freely and generously."
                   -- Benjamin Franklin

More information about the Pipet-Devel mailing list